[Pål Jensen]

Mr Benjamin barely shoots at ISOs above 400 (his words not mine). Mr Benjamin makes a living of mastering and using lighting so he doesnt have to push the shadows too far. Mr Benjamin can print billboards with the right sharpening and upsizing technique and high-tech printing facilities. Mr Benjamin has been in the business for gods know how long so we can agree he is an expert photographer. Mr Oleg and Mr Gordon discussions only affect to the remaining 99.9% of humans.

Sorry, but 99,9% of all images are shot at normal ISO values. That probably explains why we never see published the night time action shots at 6400ISO demanding ultrafast predictive AF and 6+fps.....

[bravobrown]

I'd suggest you to read this

High ISO raw comparation k20/40d/d300/d3 [Page 1]: Pentax SLR Talk Forum: Digital Photography Review

skip the usual bullshit from RMabo, read what GordonBGood writes about Oleg_V

I was not aware of that thread, however, that does not change the fact that Oleg's conclusions are suspect and arrogant. Gordon and anyone else who cares to check the math can certainly verify it from THAT year old thread. Whether or not using ONLY that method tells the entire story is another discussion. The thread you previously linked to had zero verifiable evidence to back up the asserted conclusion. Without references it's not scientific, and THAT thread from your original post is where my opinion that you disdain came from (and all other recent threads I'm aware of).

I mean that bravobrown's opinion about Oleg_V worth ZERO

I see the kindergarten class has moved here as well. Nice. You lose all credibility yourself when you go out of your way with personal attacks vis a vis RMabo, who certainly seems well informed of Pentax's company line. Whether or not he's 100% infallible doesn't give you the right to try petty baiting tactics.

[nostatic]

math != art

Sorry, but 99,9% of all images are shot at normal ISO values. That probably explains why we never see published the night time action shots at 6400ISO demanding ultrafast predictive AF and 6+fps.....

At normal iso values (100-400) in RAW, if you happen to shoot a sunset with something between you and the horizon (that makes for 90% of the landscapes that are not seascapes) so you cant use a gradual ND you still having the problem of the ugly chromatic noise on the shadows. You dont need to go 6400ISO. Obviously many people can live with strong shadows on those shoots even if you might be loosing detail that it is relevant to the shoot. There is nothing better that choose carefully for a good log or rock or flower on the foreground during a sunset to find out that if you want to show it on the image you have to blow the highlights big time.

I am not saying the sensor is a dog, I am saying that the analysis that Gordon and Oleg made affects many more pictures than people recognize without recurring to high-ISO or billboards prints.

Now if the gold standard is a 800x600 picture in PPG then I have nothing to say.

I will still paying attention to what they have to say before making a decision on which system to invest or whether or not I want to fork some extra 1200 USD on a camera which IQ doesnt better the K20d

[GaryM]

I am not saying the sensor is a dog, I am saying that the analysis that Gordon and Oleg made affects many more pictures than people recognize without recurring to high-ISO or billboards prints.

That's your opinion, and of course you are welcome to it. On the other hand, I've heard very few complaints from K20 users about shadow noise at low ISOs, in the year and a half that it's been out. In my opinion, this *problem* is 98% theoretical and 2% practical. YMMV.

[falconeye]

math != art

math is the highest art form ever created by mankind. It is pure language. Except you don't consider literature an art.

of course, literature is hard to digest if you never learned the language ...

[ManuH]

math is the highest art form ever created by mankind.

OT sorry, but math is not art. It may be beautiful, it may even elicit some emotion, but it's not art. Art is subjective, with math you cannot create 1+2=4 and make it beautiful. It's just wrong.

[bravobrown]

Sorry if I got us OT You know, math is inherently a language, the language of science... but there are certainly lots of artistic and elegant things about trying to explain the universe. That's one of the reasons lots of mathematicians become professional musicians (myself included)! Folks who are mathematically inclined are frequently proficient in languages and music. Now who created math is much deeper philosophical discussion

---

Last edited by bravobrown; 07-14-2009 at 06:51 AM.

[*isteve]

math is the highest art form ever created by mankind. It is pure language. Except you don't consider literature an art.

of course, literature is hard to digest if you never learned the language ...

Actually mathematics and language are forms of notation that allow the physical universe to be described quantitatively and qualitatively in a way thats communicable.
I actually think one is useless without the other.

The art however is in how well you express yourself, either way

But there is also a problem of "preaching to the choir". Some people can read musical notation and hear the music in their heads, like most of us can visualise the characters in a novel. But most of us prefer to hear the orchestra play music at the end of the day. With maths, most feel more comfortable with concrete examples of the outcomes.

To take one part of the image processing equation in isolation may indeed lead to a correct observation, but out of context its not very useful to a non-engineer. It would help most people to illustrate the effect using real life samples and examples that show its relevance in real life (subjective) photography in a way normal people can understand (and can replicated).

Otherwise some people inevitably read more into this than exists and make a big deal out of it, even the OP in Oleg's case. A classic case of over analysis if ever I saw one.

Similarly it would actually be more constructive for the engineering talent that abounds on the forum to suggest the best ways of minimising its effects in PP.

If we could only source some decent RAW comparison images it would be interesting to compare all the cameras in question using different RAW converters to see if subjectively the results cannot be made comparable.

[*isteve]

I was not aware of that thread, however, that does not change the fact that Oleg's conclusions are suspect and arrogant. Gordon and anyone else who cares to check the math can certainly verify it from THAT year old thread. Whether or not using ONLY that method tells the entire story is another discussion. The thread you previously linked to had zero verifiable evidence to back up the asserted conclusion. Without references it's not scientific, and THAT thread from your original post is where my opinion that you disdain came from (and all other recent threads I'm aware of).

I concur entirely with your analysis.

As for deeeejjjjthingy ignore him, most of us do. He doesnt deliver the bad news himself, he leaves that to the bears, but he's one of several carrion crows attracted to the stink and he'll attack anyone that tries to revive or bury the body once the bears have left.

[*isteve]

At normal iso values (100-400) in RAW, if you happen to shoot a sunset with something between you and the horizon (that makes for 90% of the landscapes that are not seascapes) so you cant use a gradual ND you still having the problem of the ugly chromatic noise on the shadows. You dont need to go 6400ISO. Obviously many people can live with strong shadows on those shoots even if you might be loosing detail that it is relevant to the shoot. There is nothing better that choose carefully for a good log or rock or flower on the foreground during a sunset to find out that if you want to show it on the image you have to blow the highlights big time.

I am not saying the sensor is a dog, I am saying that the analysis that Gordon and Oleg made affects many more pictures than people recognize without recurring to high-ISO or billboards prints.

I am sorry but if you try and boost the last stop of shadows by 3 EV you are simply trying to expand 1 stop of shadow detail into 4 stops when there really is not 4 stops of actual information to start with. The noise may also increase, but its not something you should really consider doing anyway because it looks so fake.

Generally expansion of the darkest 4 stops of the tone curve is what most people do to recover shadows, either by altering the tone curve or using the shadow/highlight tool. Long before noise shows up the image looks blocky and fake because the colour information is simply lacking to boost it more than a stop or so whatever camera you use, so noise levels are a theoretical issue only.

Having said that I routinely use shadow recovery on K20 files with no problems. I just dont overdo it.

For any scene with more than about 9 stops of DR, you are much better off using HDR and hey, guess what - Pentax has it built in!

If Pentax has an issue at all its simply a rather abrupt highlight rolloff, which forces you to underexpose. I would like to see that improved and from what I can see the K7 is much better.

[nostatic]

math is the highest art form ever created by mankind. It is pure language. Except you don't consider literature an art.

of course, literature is hard to digest if you never learned the language ...

I totally disagree. I see beauty and elegance in math (I've taken a fair amount), but it is not art. In fact I would consider them orthogonal though possibly complementary.

There is not such thing as "pure" anything. That is thinking in math terms

Many artistic endeavors end up being distilled down to technicalities. And at that point, it not longer is art imho. That isn't to say that artists need to be luddites - in fact I think that a good artist understands science and technology. But with math the goal is an "answer." With art there is not.

[*isteve]

I totally disagree. I see beauty and elegance in math (I've taken a fair amount), but it is not art. In fact I would consider them orthogonal though possibly complementary.

There is not such thing as "pure" anything. That is thinking in math terms

Many artistic endeavors end up being distilled down to technicalities. And at that point, it not longer is art imho. That isn't to say that artists need to be luddites - in fact I think that a good artist understands science and technology. But with math the goal is an "answer." With art there is not.

Artists and scientists have traditionally fought for the philosophical high ground, though the best of them understand that one is simply the compliment of the other. Both are necessary to completely describe our experience of the universe as human beings.

And you are right that its not necessary to define one in terms of the other. They both stand on their own merits and both have their own standards of beauty and elegance.

But applied scientists such as engineers and architectects produce artifacts for the public at large to see and use and the success of those artifacts often depends on their emotional and aesthetic qualities are often as their strength and performance. Thats why I find a purely engineering description of something as emotive as a camera rather unsatisfying.

[nostatic]

But applied scientists such as engineers and architectects produce artifacts for the public at large to see and use and the success of those artifacts often depends on their emotional and aesthetic qualities are often as their strength and performance. Thats why I find a purely engineering description of something as emotive as a camera rather unsatisfying.

This is exactly what rankles me about the decontextualized technical "reviews" on gear. People forget the point - to create photos. Not to score best on some machine test.

I have given a number of talks over the past few years where I inevitably whip one of my favorite horses: without aesthetic your technology will be lacking. You cannot leave engineers to their own devices wrt design. Similarly you can't leave the artists free reign with design. They need to inform each others. You cannot separate form from function. There is a reason for the saying, "a user interface only a programmer could love..."

[*isteve]

This is exactly what rankles me about the decontextualized technical "reviews" on gear. People forget the point - to create photos. Not to score best on some machine test.

I have given a number of talks over the past few years where I inevitably whip one of my favorite horses: without aesthetic your technology will be lacking. You cannot leave engineers to their own devices wrt design. Similarly you can't leave the artists free reign with design. They need to inform each others. You cannot separate form from function. There is a reason for the saying, "a user interface only a programmer could love..."

LOL good quote.

If you take two identical pictures and add a small amount of very fine grain noise to one of them, everyone who looks at it will think its sharper. In fact to make a very smooth noise free image look equally sharp you need to add a lot more USM. The fact is the best engineering solution is not always the nicest to look at.